DESCRIPTION: (Applicant's Abstract) The use of highly active antiretroviral therapy (HAART) has resulted in marked reductions in AIDS incidence, mortality and hospitalization. Unfortunately, these successes have not been uniform among all groups of people living with HIV, particularly among HIV infected injecting drug users (IDUs). The stunning benefits of HAART have been blunted and limited by difficulties in access and adherence to therapy. Despite the impressive successes of directly observed therapy (DOT) for tuberculosis control, only a few pilot studies of this intervention have been implemented for the treatment of HIV disease. Most of these studies of DOT using HAART have been accomplished in the controlled setting of either methadone maintenance or prisons, yet as few as 15-20 percent of IDUs are likely to be in these settings. For those who are in these settings, relapse to active drug use is common. Thus, in order to accomplish increased access to and utilization of HAART by HIV infected IDUs, innovative programs that effectively reach out-of-drug treatment IDUs are desperately needed. One such way to effectively interact with these IDUs is through needle exchange programs (NEPs). In this study, we propose to conduct a randomized, controlled trial comparing modified DOT dispensed HAART through NEP-based health services to standard of care. Recruited subjects of out-of-drug treatment HIV+ IDUs will be randomized based on census-block code and experience with antiretroviral therapy. Modified DOT will be administered to the intervention group; for twice daily regimens, they will receive one dose per day as DOT and will receive a programmed reminder to take their next dose on their own. Primary endpoints include standardized measures of adherence, reduction in HIV-1 RNA levels and time to development of a primary HIV-1 resistance mutation. Secondary measures of interest include quality of life measures, health care utilization and entry into drug treatment.